Area-Efficient Fault Detection Mechanism for Carry-Lookahead and Carry-Save Adders
SilpaKesav Velagaleti, Gujjari Hruthik · 2024
In the inherent self-checking capabilities of an adder involve real-time error detection and correction during addition operations. This research paper introduces a novel approach to the self-checking attributes of both Carry-Lookahead Adder (CLA) and Carry-Save Adder (CSA) by incorporating distributed fault detection mechanisms. The checking process employs a hot-standby methodology coupled with partial reconfiguration, enabling the replacement of malfunctioning modules during runtime. This methodology minimizes area overhead for the self-recovery process by selectively addressing faulty modules instead of overhauling the entire system. The proposed self-checking CSA exhibits a 12% reduction in area while maintaining equivalent performance. On the other hand, as compared to the traditional CLA design, the self-checking CLA suffers an area overhead of 161.5% due to its high fault coverage. However, this fault coverage represents a 35.3% reduction when contrasted with the partial reconfiguration CLA. In addition, compared to the traditional ripple carry adder, the suggested 64-bit CLA has a 40.7% faster in latency.